Detection of SARS-CoV-2 from raw patient samples by coupled high temperature reverse transcription and amplification

Abstract

SARS-CoV-2 is spreading globally with unprecedented consequences for modern societies. The early detection of infected individuals is a pre-requisite to contain the virus. Currently, purification of RNA from patient samples followed by RT-PCR is the gold standard to assess the presence of this single-strand RNA virus. However, these procedures are time consuming, require continuous supply of specialized reagents, and are prohibitively expensive in resource-poor settings. Here, we report an improved nucleic-acid-based approach to detect SARS-CoV-2 with the ability to detect as little as five viral genome equivalents. The approach delivers results without the need to purify RNA, reduces handling steps, minimizes costs, and allows evaluation by non-specialized equipment. The use of unprocessed swap samples is enabled by employing a heat-stable RNA- and DNA-dependent DNA polymerase, which performs the double task of stringent reverse transcription of RNA at elevated temperatures as well as PCR amplification of a SARS-CoV-2 specific target gene. As results are obtained within 2 hours and can be read-out by a hand-held LED-screen, this novel protocol will be of particular importance for large-scale virus surveillance in economically constrained settings.

Fig 1. A RNA- and DNA-reading heat-stable polymerase reverse transcribes and amplifies viral RNA.

A) Schematic overview of the SARS-CoV-2 genome. The target sequences for the N1 primers and probe are marked in red and green, respectively. The R2 primer binding sequence is underlined. Sequences divergence between SARS-CoV-2 and SARS-CoV-1 genomes are highlighted in blue. B) Performance of Volcano3G polymerase was compared to Taq polymerase using plasmid DNA or in vitro transcribed RNA as template (5000 viral genome equivalents). C) Determination of the linear dynamic range for the Volcano3G protocol with or without an additional primer (R2) for optimized reverse transcription at a final concentration of 250 nM. In vitro transcribed RNA containing the Sars-CoV-2 N amplicon was serially diluted in the range from 1x10⁶ copies to 10 copies. D) Limit of detection (LOD) was assessed with serial dilutions ranging from 20 to 1 copy per reaction (n = 6 for each dilution). The fraction of positive reactions (y-axis) were plotted against the log-transformed number of RNA copies per reaction. Addition of R2 primer enhances the performance at lower copy-numbers. E) Amplification curves showing the performance of Volcano3G on isolated RNA from two COVID-19 patients in presence or absence of R2.

Fig 2. SARS-CoV-2 detection by high-temperature RT-PCR in a patient cohort delivers results consistent with the standard procedure.

A) RNA was isolated from nasopharyngeal- and throat swab samples (n = 43) and SARS-CoV-2 and RNAseP were detected using the Volcano3G protocol. N1 amplicon (blue), RNaseP gene (gray). Water was used as a non-template control (light gray). B) Identical samples were processed in parallel in an accredited diagnostic lab using the Allplex 2019-nCoV assay from Seegene. Direct comparison of assay results reveals 100% concordance of Volcano3G with the reference assay. C) Cq values obtained with Volcano3G were lower than those obtained with the reference assay (ΔCq = 6.4 +/- 0.78). D) For each positive patient sample, the Cq values obtained with both assays were plotted against each other for linear regression analysis. A highly significant correlation of Volcano3G with the reference assay was observed (r² = 0.98, p<0.0001).

Fig 3. High-temperature RT-PCR using Volcano3G polymerase allows SARS-CoV-2 detection from unprocessed patient samples.

A) Nasopharyngeal- and throat swab samples (prepared in water) were added directly as template for RT-qPCR using the Volcano3G protocol. Representative amplification curves of patients with high (dark blue), medium (medium blue) and low Cq as well as negative patients (light blue) are shown. B) RNA was isolated from the remaining patient material and analysed in an accredited diagnostic lab using the Allplex 2019-nCoV assay from Seegene. The samples were divided in 3 groups based on cq (<24, 24–30 and >30) and each sample was analysed repeatedly (3–4 times) by high-temperature RT-PCR. The bar diagram depicts the average frequency of detection (± SEM) for each group (patient samples in each group: cq < 24 n = 12; cq 24–29 n = 12; cq > 30 n = 9 and negative samples n = 12). C) The cq values of each patient sample are compared between the reference protocol and the Volcano3G direct approach. Dotted red line indicates the cut-off, were the direct assay looses sensitivity. D) For each positive patient sample, the Cq values obtained in 3–4 repetitions with the high-temperature RT-PCR (Volcano3G direct input; mean +/- sd) were plotted against the cq-values obtained with the standard RT-PCR on isolated RNA for linear regression analysis (r² = 0.779, p<0.0001). E) RT-PCR analysis of 100 copies of in vitro transcribed RNA spiked with varying amounts of pooled patient material from 5 confirmed negative patients. F) 10 swabs were eluted a second time according to our optimized protocol (low elution volume, protease K treatment and additional MgCl₂). In addition, 10 confirmed negative swab samples were included. G) 5 Volcano3G reactions from the cohort of unprocessed samples presented in Fig 3A (in low-translucent white tubes) were photographed on a blue light transilluminator. Reference Cq values (Seegene Allplex on purified RNA) and Cq values obtained with Volcano3G using unprocessed samples (measured Cq) are depicted above each tube.